FIGURE CAPTIONS
Figure 1. Oblique helicopter photograph of Whoopup Canyon, Wyoming, after the fire.
Sandstone joint faces on the canyon walls host petroglyph panels.
Figure 2. Optical thin section view of ash adhererance to spalled surfaces.
A Sample WC-03-43 derives from Panel 3.1, a fire spall where the pre-existing
weathering rind formed the detachment surface. The ash survived two years on the panel
face by mixing with fragments of weathered quartz. The granular appearance of the
quartz can be seen in situ within in tact weathering rinds.
B. Sample WC-03-06 is from Fat Man With Staff petroglyph. The pre-fire varnish
was much thicker with a complete microlamination sequence, but the fire spalled most of
the varnish. The varnish detached in zones of porosity weakened by cation leaching.
Then, post-fire soil erosion washed sediment over the petroglyph, and a mixture of
potassium-rich ash and inorganic sediment now adsorbs onto the eroded-varnish surface.
Figure 3. Optical thin section image of chalk applied to petroglyph surfaces survived the
fire. Sample WC-03-30, from a Round-Bodied Human petroglyph, reveals that chalk
adhered to the surface of the varnish so strongly that it took erosion of the entire sequence
of rock and varnish to remove this foreign material. According to Liu's (2003)
calibration, this section displays a Holocene (only orange-yellow) varnish
microlamination sequence.
Figure 4. The wildfire appears to have caused, in some cases, the varnish to separate from
the underlying weathering rind. In these optical thin section examples, erosion of the preweathered rock resulted in a pattern where remnants of varnish exist only in tiny pockets.
Arrows point to quartz that eroded on surfaces decayed over a long period of time, well
prior to the fire event. Mechanical fire weathering, in contrast, generates much cleaner
fractures (Dorn, 2003). The scale bar applies to all photomicrographs, where the
petroglyph samples from top to bottom are WC-03-30 (Round Bodied Human), WC-0332 (Cat), and WC-03-15 (Ball-Footed Animal).
Figure 5. Optical thin section sample WC-03-12 from a panel that acted like a fire
chimney. Detachment surfaces are in the weathering rind. Two potential consequences of
the fire can be seen. First, the host quartz appears to have undergone thermal fracturing.
The linear fracturing, exemplified by the arrows, appears to cut across both relatively
fresh and more weathered grains. Certainly, we saw grain fractures in the 1991 pre-fire
samples. However, the abundance of these subparallel fractures appears to have
increased dramatically. Second, rock varnish on the panel appears to have experienced
spalling.
Figure 6.. Comparison of (A) 1991 pre-fire and (B) 2003 post-fire samples of a
"hobbled deer" petroglyph (pre-fire sample WC-91-16 at panel 24.1; post-fire sample
WC-03-16). The same types of crystal defects and decrystallization occur in pre-fire and
post-fire imagery. These HRTEM views also illustrate that Whoopup varnish starts to
form on quartz surfaces with an irregular topography — helping in varnish nucleation
(Dorn, 1988).
Figure 7. This panel (20.2) BSE sample WC-03-11 came from an unexposed joint face,
approximately 40 centimeters up into a joint still covered by an overlying block of
sandstone. The sample displays the weakest of the case hardening cements, a relatively
porous silica glaze. Composed of mostly silicon, with minor amounts of aluminum and
potassium, this silica glaze is typical of coatings found on the walls of sandstone joint
faces at Whoopup. This "inherited" silica glaze displays two textures, disoriented
platelets (left) and granular forms (right). Left scale bar is 210 µm across and the right is
260 µm across.
Figure 8. These BSE images from a Big Elk Petroglyph (WC-03-34, WC-91-38, Panel
6.2) focus on case hardening by a mixture of silica glaze and rock varnish.
Images A1, B1/B2, and C1/C2 all show case hardening of the petroglyph’s quartz by
a mix of silica glaze and rock varnish, where the silica could be "inherited from
subsurface cementation (Fig. 7). Rock varnish would then be added after the panel face
is exposed by spalling.
Images A2 and D1/D2 illustrate how the fire and microenvironment complicates
varnish dating of petroglyphs. Image A2 displays a covering of clays and ash adhered
reasonably well to the varnish. This ash covering contains a spike in potassium (K). The
topographic (secondary) image D1 highlights the presence of lichens growing on top of
the varnish. The chemical (backscatter) D2 image shows erosion and enhanced cation
leaching of the rock varnish — resulting in a much lower cation ratio than varnishes not
influenced by lichens.
Image widths: A1 (600 µm); A2 (140 µm); B1 (560 µm); B2 (180 µm); C1 (360
µm); C2 (110 µm); D1 (330 µm); D2 (330 µm). Image A2 is a close-up off the letter
“C” in image A1.
Figure 9. BSE samples from the Human petroglyph (Panel 30.1, WC-03-49m, WC-9138), exemplify case-hardening agent by iron film. Image A1 displays the full stratigraphy
from top to bottom of: surficial layer of rock varnish; case hardening by a mixture of
silica glaze and rock varnish; case hardening by iron film (a water-flow deposit
containing silicon and iron); and a weathering rind underneath the iron film characterized
by decayed quartz.
Images A2 and A3 show progressive close-ups of the top three layers. Note the
greater porosity of the varnish/Si-glaze case hardening cement, compared with the iron
film. Also note how the iron film is thickest towards the top and thins progressively
further away from the panel surface. Image C3 presents the reasonably-distinct contact
between these two case hardening zones.
Images C1 and C2 present the prevalent condition at the base of the samples: thorough
decay of the quartz minerals. The quartz grains appear to be dissolving and undergoing
granular decay. Such a condition could occur if acidic, iron-rich waters were dammed
behind the iron film.
Image B2 (and the lower magnification B1) illustrates a relatively common
correlation: where lichens appear, the varnish appears to be completely eroded or in a
state of erosion. The acid released by lichens dissolves the manganese and iron “glue”
that holds the varnish together — resulting in the loss of varnish. The source of the iron
could derive, in part, from lichen-generated acidity.
Image widths: A1 (2200 µm); A2 (1400 µm); A3 (530 µm); B1 (240 µm); B2 (150
µm); C1 (250 µm); C2 (200 µm); C3 (600 µm).
Figure 10. These BSE images from a Dog petroglyph (Panel 6.2, WC-03-37, WC-91-39)
illustrate where case hardening by iron films is right next to the surface. Images A, B, C,
and G show iron film, alone, case hardening the sandstone. The varnish/Si-glaze case
hardening appears to be missing, perhaps pecked away during petroglyph manufacturing.
The rock varnish can be seen accreting directly on top of the iron film in images B, C and
G. The iron film is thickest at the top and thins into micron threads at a depth of a few
millimeters. Images D, E and F display mechanical fracturing of the quartz grains by the
precipitating iron films. Image widths: A (2200 µm); B (110 µm); C (270 µm); D (100
µm); E (100 µm); F (110 µm).
Figure 11. From WC-03-43 (panel 3.1), these BSE images of spalled fragments illustrate
the typical pattern of detachment in the weathering rind. Image A maintains a sequence
of rock varnish underlain by quartz grains cemented by a combination of silica glaze and
rock varnish. Image B presents a higher magnification view of the varnish/Si-glaze
mixture case hardening the panel. Images C, D, E and F display views of the underside
of spalls; all detachments took place along chemically-weathered surfaces. Image
widths: A (1800 µm); B (140 µm); C (180 µm); D (140 µm); E (150 µm); F (150 µm)
FIGURES
Figure 1.
Figure 2.
Figure 3
Figure 4.
Figure 5.
Figure 6.
Figure 7.
Figure 8.
Figure 9.
Figure 10.
Figure 11.